Catching up with Krugman

Here’s Paul Krugman on Finland:

Why can’t Finland recover this time? Debt is not a problem; borrowing costs are very low. But it’s all about the euro straitjacket. In 1990 the country could and did devalue, achieving a rapid gain in competitiveness. This time not, so that there is no quick way to adjust to adverse shocks:


This shouldn’t come as a surprise; it’s the core of the classic Milton Friedman argument for flexible exchange rates, and in turn for the tradeoff at the core of optimum currency area theory. The trouble in Finland is what everyone expected to go wrong with the euro.

What’s going on in Greece represents a whole additional level of hurt, which nobody saw coming. But it’s important, I think, to realize that even countries that didn’t borrow a lot, didn’t experience large capital inflows, basically did nothing wrong by the official criteria, are nonetheless suffering in a major way.

Krugman’s right that for the most part  (excluding Greece) the euro is failing in exactly the way people like Krugman and Friedman expected it to fail.

Here’s an interesting Krugman observation on Grexit:

But the bigger question is what happens a year or two after Grexit, where the real risk to the euro is not that Greece will fail but that it will succeed. Suppose that a greatly devalued new drachma brings a flood of British beer-drinkers to the Ionian Sea, and Greece starts to recover. This would greatly encourage challengers to austerity and internal devaluation elsewhere.

This is the problem the gold bloc faced after Britain left gold in 1931.  I have a similar view, although I’m a bit less worried about Greece leaving the euro than Krugman, but partly because I think an “experiment” would be useful.  Unfortunately the new Greek government is so incompetent I’m not even sure that devaluation would help, but it’s worth a shot.  (Just to be clear, at this late date I believe Greece would be much better off doing structural reforms and staying within the euro.)

Here’s Krugman (implicitly) discussing the never reason from a price change problem:

In the left panel, you see the Fed funds rate seesawing as the Fed tried to grapple with inflation — and you see housing starts move strongly in the opposite direction, plunging when rates rose and soaring when they fell. In the right panel, however, you see housing starts and interest rates moving in the same direction — plunging together. So is there no relationship?

Bad answer. The situations are different in a fundamental way. In the 80s interest rates were being driven by fear of inflation; from the point of view of the housing market, they were more or less endogenous. In the post-2006 period they were being driven largely by the housing bust itself. So the 80s experience was a sort of natural experiment in the effects of rate changes, whereas more recent events are an illustration of reverse causation.

And I’ve been arguing for a long time that there was a regime shift in the late 1980s, that as inflation fears were replaced by the Great Moderation, we entered an era of postmodern recessions in which monetary policy was trying to clean up after bubbles rather than curb inflation.

The point, then, is that when you look for the effects of monetary policy, it’s often important to distinguish between eras when interest rates are a cause and eras when they’re an effect — and it’s not that hard to know which eras we’re talking about.

My only quibble is that instead of saying it’s only OK to reason from a price change in certain occasions, I prefer saying never reason from a price change—always reason from the thing that caused the price to change (such as tight money in 1981.)  Tight money depressed housing in 1981, no need to even mention interest rates.

PS.  I have a new post on current Fed policy over at Econlog.

Marcus Nunes was right

More than three years ago Marcus Nunes did a post taking on two of my favorite macroeconomists, Robert Hall and Greg Mankiw.  It hardly seems like a fair fight, as Marcus is merely an amateur market monetarist down in Brazil.  And yet, I’ll show that Marcus was correct in both disputes.  In a January 2012 post, Marcus quotes from a Ryan Avent post.  Here’s Ryan:

This time around, Mr Hall addressed the point head on. He noted that in a liquidity trap, the real rate of interest was simply equal to the negative inflation rate. In other words, if the Fed’s nominal rate is at 0% and the inflation rate is 2%, then the real rate of interest is -2%. If a -3% real interest rate is necessary to clear the economy, then all that’s needed is a higher rate of inflation—3% rather than 2%. Mr Hall noted that this was an important point because potentially the Fed could have an enormously helpful impact on the economy simply by raising inflation just a little. And here’s where things got topsy-turvy. Mr Hall argued that (my bold):

  1. A little more inflation would have a hugely beneficial impact on labour markets,
  2. And a reasonable central bank would therefore generate more inflation,
  3. And the Federal Reserve as currently constituted is, in his estimation, very reasonable; therefore
  4. The Federal Reserve must not be able to influence the inflation rate.

Both Ryan Avent and Marcus Nunes thought Hall was wrong.  So did I.  I could have given dozens of reasons.  Bernanke said the Fed could do more.  The Fed had not done the things Bernanke recommended the Japanese do.  The market response to QE rumors. The difference between central banks that did QE and those that didn’t.  Since this time we have lots more evidence, such as the Abe policy in Japan, which has pushed the yen from 80 to 125/dollar.  Despite the recent dip due to falling oil prices, Japanese inflation since 2013 is significantly higher than before.

But even if that was all wrong, we can still be 100% certain that Hall was completely incorrect.  Indeed it’s not even a debatable point anymore.  Why?  I hope it’s obvious to everyone by now.  Hall is wrong because the Fed plans to raise interest rates later this year, even though they expect inflation to continue running below 2%.  So the Fed does not want to raise its inflation target to 3%.  Period. End of debate.

[Isn’t it wonderful when economic debates get resolved with 100% certainty!  It happens so rarely.]

Now on to the next victim, Greg Mankiw.  Here he quotes Mankiw discussing the condition of the economy in early 2012, in light on Mankiw’s own version of the Taylor Rule:

Not surprisingly, the rule recommended a deeply negative federal funds rate during the recent severe recession.  Of course, that is impossible, which is why the Fed took various extraordinary steps to get the economy going.  But note that the rule is now moving back toward zero.  As Eddy points out, “At the current inflation rate, the unemployment rate needs to drop to 8.3% from the current 8.5% for the model to signal positive rates. We’re getting close.”

To be fair to Mankiw, he doesn’t explicitly say the Fed should raise rates when unemployment falls to the low 8% range, even if inflation stayed about the same. There’s some wiggle room.  So let me just say that I think the vast majority of readers would have assumed that Mankiw was stating that policy implication with approval.

In any case, we now know that tightening monetary policy in early 2012 would have been a big mistake.  Actual unemployment has fallen rapidly to 5.4%, and inflation remains well below the 2% target.  Furthermore, inflation is expected to remain below target for an extended period, while unemployment is expected to keep falling. Tightening policy in 2012 would have been a mistake, even if you were a hawkish Fed policymaker who would prefer to ignore unemployment and focus like a laser on their 2% inflation target.

In contrast, the ECB did raise rates a few months before the Nunes post, and we now know that the result was disastrous.

The bottom line here is that the Taylor Rule (in any form) is not a reliable instrument rule.  That’s not to say that the Taylor Principle is without value—I think it contributed to the successful Fed policy of the Great Moderation period.  But as a rigid instrument rule it simple doesn’t work.

PS.  In my tradition of “burying the lede”, let me point to a comment left by Julius Probst:

Yesterday, Larry Summers held a speech at DIW Berlin here in Germany, talking about secular stagnation.
He mentioned that bond markets expect low trend real growth rate in the years to come, low inflation rates and low interest rates – all of that is true in my opinion and I think that the FED is much too optimistic expecting long-term interest rates at 4% in a few years. It won’t happen.

I got to ask Larry Summers a question whether he favors more aggressive monetary stimulus – he does – and whether he favors a different monetary target.

He was against a 4% inflation target, but he favored a NDGP target !!!! I think that might have been the first time he did so in public ? At least the first time I know about it.

The same with me, I’d never heard that he favored NGDP targeting.  Maybe I backed the wrong horse in the Fed race last year.  Oh well.  I’d appreciate it if someone could send a link as soon as possible.  With endorsements by Woodford, Romer, McCallum, Frankel, DeLong, Krugman (sort of), Bullard, and now Summers, there is obviously strong momentum toward NGDP targeting among the macro elite.  So much for the silly view that MM ideas are becoming less popular.  (Of course I’m not claiming that our blog posts actually caused any of these conversions, just that our policy ideas are getting more popular.)

When did the left jump the shark?

Was it when they started talking about helicopter drops? Or perhaps the point where the LA labor unions asked to be exempted from the new minimum wage law?  Maybe, but I vote for this gem from The Guardian, which (believe it or not) is actually considered one of the UK’s respectable newspapers:

Landlords are allowed to deduct a wide range of expenses, on top of mortgage interest costs, before they have to pay tax on their rental income. These allowable expenses include the cost of insurance, maintenance and repairs, utility bills, cleaning and gardening, and legal fees. Ordinary homeowners are not entitled to similar privileges.

Tim Worstall and Britmouse beat me to it.

People sometimes ask me why I consider The Economist to be the best newspaper. Here’s one reason—can you even imagine a paragraph that idiotic in The Economist?

PS.  Full disclosure; I’m one of those “privileged” landlords.

The perfect villain

Poor Dennis Hastert.  He picked the wrong country to get born into.

1.  He picked a Puritan state in a Puritan country with a higher age of consent than any European country, save Ireland, Cyprus and Turkey.

Screen Shot 2015-05-31 at 9.39.44 AM

2.  He picked a country where (according to Politico) the GOP cares so little for the lives of blacks and gays that it won’t lift a finger to stop a needle/HIV epidemic until it starts hitting Red State voters.

3.  He picked a country where the Dems think it’s a crime to frequently withdraw $5000 in cash from your own bank account, and use the cash for perfectly legal activities.

4.  He picked a country where voters have so much faith in law enforcement that they make it a crime to lie to police, even to cover up an embarrassing personal scandal.

5.  He picked a country where people are obsessed over any sex where there is a “power imbalance.”

PS.  Attention commenters; I’m offering no editorial comment, just describing things as they are.  If you don’t like the post, don’t blame me, change America.

Bullard, et al, on NGDP targeting

Lots of people (including James Bullard) have sent me a new paper by Costas Azariadis, James Bullard, Aarti Singh, and Jacek Suda.  Here is the abstract:

We study optimal monetary policy at the zero lower bound. The macroeconomy we study has considerable income inequality which gives rise to a large private sector credit market. Households participating in this market use non-state contingent nominal contracts (NSCNC). A second, small group of households only uses cash and cannot participate in the credit market. The monetary authority supplies currency to cash-using households in a way that changes the price level to provide for optimal risk-sharing in the private credit market and thus to overcome the NSCNC friction. For sufficiently large and persistent negative shocks the zero lower bound on nominal interest rates may threaten to bind. The monetary authority may credibly promise to increase the price level in this situation to maintain a smoothly functioning (complete) credit market. The optimal monetary policy in this model can be broadly viewed as a version of nominal GDP targeting.   (emphasis added)

Obviously I like this paper, even though they rely on nominal debt contracts rather than my preferred sticky nominal wages as the key friction.

In their model, when the zero bound threatens to appear, policymakers deviate from standard policy by engineering a rise in the price level.  This sounds vaguely NeoFisherian:

The price level approach involves a promise to engineer an increase in the price level one period in the future sufficient to keep the net nominal interest rate positive. This promise is sufficient to ensure that the net nominal interest rate remains positive and the complete credit market policy remains intact.

As I indicated earlier, the NeoFisherians have a point when they suggest that raising interest rates can raise inflation.  Their claim is actually wrong in the context of current Fed operating procedures, but would be correct under certain policy regimes.  Notice that in this example, monetary policy raises interest rates and expected inflation together.

Continuing on:

However, this policy has a drawback: The price level policy harms cash-using households relative to the policy away from the zero lower bound. As additional shocks hit the economy, the zero lower bound situation will eventually dissipate and special policy actions will prove temporary.

We conclude that in economies where the key friction is NSCNC and the net nominal interest rate threatens to encounter the zero lower bound, monetary policymakers may wish to respond with a price level increase. A chief rival to this response observed in actual economies—forward guidance on the length of time the economy will remain at the zero lower bound beyond the time when that bound is actually binding—would be inappropriate in the theory presented here.

I absolutely LOVE this passage.  I’ve repeatedly criticized the New Keynesian view that the way out of a liquidity trap is to keep interest rates low for an extended period. Rather I’ve suggested using NGDPLT to keep interest rates from falling to zero in the first place.  I’ve argued that the “zero rates for an extended period” approach is hard to distinguish from Japanese policy, and that the only country with sufficiently expansionary policy during the Global Financial Crisis was Australia, which never hit the zero bound.  Australia had 6.5% NGDP growth from 1996 to 2006, and then 6.5% NGDP growth from 2006, to 2012.

As far as currency holders being hurt:

1.  They’d be helped far more by the positive employment affects of effective monetary policy.

2. If we care so much about people who choose to conduct their affairs in currency, then why was Dennis Hastert arrested?  Alternatively, if the government assumes that people dealing in cash are criminals, it seems plausible that the optimal rate of tax on currency would be even higher than implied by the occasional one time price level increases in this proposed regime.  (Yes, I’m being sarcastic.)

Eggertsson and Mehrotra (2014) follow authors like Benhabib, Schmitt-Grohe, and Uribe (2001), Bullard (2010), and Caballero and Farhi (2015) in modeling the zero lower bound as at least potentially a permanent outcome. In the present paper, the zero lower bound can be encountered because of large and persistent aggregate shocks, but is ultimately temporary.

Of course there is always a trend rate of inflation/NGDP growth sufficiently high to prevent the zero bound from being permanent.

How is it that the monetary policymaker can control the price level in this model? The policymaker supplies currency, H (t); to the non-participant households—the cash users.

It warms my heart to see people talk about controlling the price level by controlling the currency stock.  Eugene Fama must be smiling somewhere.

The bottom line is that although the policy implications are discussed in terms of price level adjustments, the policy implications are very close to NGDP targeting:

In terms of inflation rates, inflation would be relatively high at times when output is growing slowly and inflation would be relatively low when output is growing rapidly. On average, the net inflation rate would be zero (which we have defined as the inflation target here), and the policymaker would achieve the targeted rate of inflation in an average sense. It is the nature of the reaction to shocks which distinguishes the complete markets policy rule from the price stability rule, not the average rate of inflation.

.  .  .

Another way to view the optimal monetary policy in the low volatility economy is as nominal income targeting.

They cite previous research by Kevin Sheedy (2014) and Evan Koenig (2013), which reached similar conclusions.  If one were to add the sticky wage friction, I believe the argument for NGDP targeting would become even stronger.

Again, the key insight here is that you want a policy regime that prevents nominal interest rates from falling to zero.  In my view NGDPLT is the simplest and most politically feasible way of doing so.

Lots of catching up to do, I’ll get to the older comments eventually.

PS.  The Wall Street Journal indicated that Bullard made some additional comments about the research, which I discuss over at Econlog:

He said his model is consistent with the Fed maintaining its 2% inflation target, but would simply mean the central bank would work more aggressively to counter any undershooting or overshooting of that objective. For instance, a nominal GDP target would have likely prescribed even more aggressive Fed monetary easing during the crisis-ridden 2007-2009 period, Mr. Bullard said.

“If you took this paper seriously we should have done even more in 2007-2009,” he said. “Something like nominal GDP targeting, if it’s appropriately formulated, does look like optimal policy.”

:)   :)   :)   :)  :)   :)   :)   :)   :)   :)   :)